INDEX

Explanations

terms related to potential occurrences or situations that may require specific actions or preparations

oai_token-act-pair · gpt-3.5-turbo

references to specific events or occurrences

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

 Flavoring

-0.78

zie

-0.76

Rated

-0.68

 Views

-0.65

MET

-0.63

pees

-0.61

Wow

-0.59

 Moves

-0.59

 Bleach

-0.58

 Ammunition

-0.57

POSITIVE LOGITS

ual

0.74

uner

0.71

ually

0.69

igation

0.68

uate

0.67

amination

0.67

 unsatisf

0.66

uating

0.65

uates

0.63

 circumstances

0.63

Activations Density 0.023%

terms related to potential occurrences or situations that may require specific actions or preparations

references to specific events or occurrences

No Comments

No Known Activations