INDEX

Explanations

relevant actions or events described in past tense

oai_token-act-pair · gpt-3.5-turbo

occurrences of notable events or significant actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 endif

-0.69

OUND

-0.56

 depending

-0.52

Magikarp

-0.52

%.

-0.51

ailable

-0.51

ounding

-0.51

 subdu

-0.50

%);

-0.50

obin

-0.50

POSITIVE LOGITS

 yesterday

0.67

©¶æ

0.65

dit

0.63

 last

0.60

terday

0.59

Merit

0.57

ruary

0.57

?,

0.55

>>>>>>>>

0.55

cember

0.54

Activations Density 0.651%

relevant actions or events described in past tense

occurrences of notable events or significant actions

No Comments

No Known Activations