INDEX

Explanations

phrases indicating future actions or plans

oai_token-act-pair · gpt-3.5-turbo

repeated phrases indicating refusal or denial

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

cius

-0.77

LAB

-0.65

 ancest

-0.63

hiba

-0.62

eg

-0.61

 Enhancement

-0.58

Stre

-0.57

aan

-0.57

Condition

-0.57

uden

-0.56

POSITIVE LOGITS

 anymore

1.48

 anywhere

1.07

nor

1.06

 bother

0.98

 anybody

0.93

 anything

0.92

 whatsoever

0.92

any

0.90

 slightest

0.88

 necessarily

0.88

Activations Density 0.079%

phrases indicating future actions or plans

repeated phrases indicating refusal or denial

No Comments

No Known Activations