INDEX

Explanations

descriptions or mentions of geographical locations

oai_token-act-pair · gpt-3.5-turbo

references to global news or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

ample

-0.70

attribute

-0.65

nesday

-0.65

hound

-0.64

ATA

-0.64

 Bundy

-0.64

mare

-0.62

avorite

-0.61

Rule

-0.61

uations

-0.60

POSITIVE LOGITS

 sources

0.84

 standpoint

0.76

 perspectives

0.75

 Vend

0.74

ablishment

0.72

 vantage

0.72

 perspective

0.71

ville

0.70

 rooft

0.65

ources

0.65

Activations Density 0.088%

descriptions or mentions of geographical locations

references to global news or events

No Comments

No Known Activations