INDEX

Explanations

words related to specific names, titles, or labels within a context that includes elements of intensity or importance

oai_token-act-pair · gpt-3.5-turbo

references to specific media titles or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

ulates

-0.80

perature

-0.80

iage

-0.79

yright

-0.78

etheless

-0.77

abases

-0.73

isations

-0.73

etooth

-0.72

oshop

-0.72

ounter

-0.71

POSITIVE LOGITS

 Breaker

1.06

 Reborn

1.04

 Brewing

0.91

 Collection

0.91

 Series

0.89

 Rising

0.89

 Lounge

0.89

 Runner

0.88

 Zone

0.87

 Collective

0.86

Activations Density 0.223%

words related to specific names, titles, or labels within a context that includes elements of intensity or importance

references to specific media titles or entities

No Comments

No Known Activations