INDEX

Explanations

references to probabilities or likelihoods

oai_token-act-pair · gpt-3.5-turbo

phrases that describe likelihood or probability

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

DATA

-0.83

Meta

-0.81

ILE

-0.78

Series

-0.73

Physical

-0.71

Hardware

-0.69

Nap

-0.69

Switch

-0.68

CAST

-0.68

Cart

-0.68

POSITIVE LOGITS

pring

0.92

llor

0.91

 chances

0.87

hift

0.87

ensical

0.75

 Rouhani

0.72

awaru

0.68

 abound

0.68

 worsh

0.67

¬¼

0.67

Activations Density 0.019%

references to probabilities or likelihoods

phrases that describe likelihood or probability

No Comments

No Known Activations