INDEX

Explanations

mentions of technology, possibly related to information technology or communication

oai_token-act-pair · gpt-3.5-turbo

words that are part of varied classifications or categories, likely with a focus on organizations and structures

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 destro

-0.81

 sacrific

-0.80

ãĥ¯ãĥ³

-0.74

 tremend

-0.73

 unden

-0.73

 challeng

-0.70

 conclud

-0.69

Ò

-0.68

 corrid

-0.68

EStreamFrame

-0.68

POSITIVE LOGITS

 Profile

0.87

 Seeds

0.69

 Gardens

0.68

 Edit

0.68

 Inher

0.68

 Emblem

0.67

 Emails

0.64

 Rating

0.64

ionage

0.63

achus

0.62

Activations Density 0.378%

mentions of technology, possibly related to information technology or communication

words that are part of varied classifications or categories, likely with a focus on organizations and structures

No Comments

No Known Activations