INDEX

Explanations

phrases indicating the lack of response or acknowledgment to a certain situation

oai_token-act-pair · gpt-3.5-turbo

negations or expressions of non-responsiveness in communication

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

interstitial

-0.79

entric

-0.73

stakes

-0.73

FAQ

-0.70

ership

-0.70

Reader

-0.70

wrapper

-0.70

Maker

-0.67

former

-0.66

=-=-=-=-=-=-=-=-

-0.65

POSITIVE LOGITS

 necessarily

1.04

 hesitate

0.98

icably

0.91

 adequately

0.90

 initially

0.88

 originally

0.88

 exactly

0.87

 foresee

0.86

 disappoint

0.86

 bother

0.81

Activations Density 0.151%

phrases indicating the lack of response or acknowledgment to a certain situation

negations or expressions of non-responsiveness in communication

No Comments

No Known Activations