INDEX

Explanations

terms related to drawing activities

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "drawn" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

mys

-0.66

chuk

-0.64

estine

-0.64

rella

-0.63

Dol

-0.62

via

-0.62

 Guys

-0.62

 docs

-0.60

aband

-0.60

 Markets

-0.59

POSITIVE LOGITS

 drawn

3.92

drawn

2.33

 Drawn

2.11

 draw

1.71

 drawing

1.64

 draws

1.51

 drew

1.51

draw

1.49

 attracted

1.37

 Draw

1.33

Activations Density 0.019%

terms related to drawing activities

occurrences of the word "drawn" in various contexts

No Comments

No Known Activations