INDEX

Explanations

phrases describing individuals who perform specific actions or hold certain roles

oai_token-act-pair · gpt-3.5-turbo

references to individuals who perform notable actions or roles

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

ply

-0.78

Bron

-0.72

fml

-0.71

van

-0.68

early

-0.66

aned

-0.63

enei

-0.61

ount

-0.61

backs

-0.61

vette

-0.61

POSITIVE LOGITS

 Teresa

0.71

 manufact

0.69

iest

0.68

 Gamergate

0.68

 tallest

0.67

IG

0.67

 Tata

0.66

 hardest

0.66

 Restore

0.65

Alc

0.65

Activations Density 0.376%

phrases describing individuals who perform specific actions or hold certain roles

references to individuals who perform notable actions or roles

No Comments

No Known Activations