INDEX

Explanations

references to profiles or profile-related actions

oai_token-act-pair · gpt-3.5-turbo

references to user profiles

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

actic

-0.77

nuts

-0.70

hers

-0.69

 relent

-0.69

mental

-0.66

AFTA

-0.66

RECT

-0.65

shall

-0.65

meat

-0.64

VEN

-0.64

POSITIVE LOGITS

 profiles

1.08

 profile

1.03

ocl

0.88

onym

0.86

ographies

0.84

 picture

0.74

ography

0.72

Profile

0.69

 Detail

0.69

ographers

0.69

Activations Density 0.044%

references to profiles or profile-related actions

references to user profiles

No Comments

No Known Activations