INDEX

Explanations

expressions of value or merit associated with people, ideas, or actions

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

words related to being deserving or having value

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

zyme

-0.66

anwhile

-0.62

 Schwarz

-0.62

è¦ļéĨĴ

-0.62

-+-+-+-+

-0.62

cation

-0.61

many

-0.61

Phase

-0.61

rief

-0.61

acid

-0.61

POSITIVE LOGITS

 worthy

0.98

 deserving

0.91

nesses

0.81

 successors

0.80

minded

0.78

 contenders

0.78

NESS

0.78

 successor

0.78

 unworthy

0.74

 newcom

0.74

Activations Density 0.009%

expressions of value or merit associated with people, ideas, or actions

words related to being deserving or having value

No Comments

No Known Activations