INDEX

Explanations

phrases related to uncertainty or decision-making processes

oai_token-act-pair · gpt-3.5-turbo

phrases that express being "in" or "out" of various situations or contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ector

-0.69

elf

-0.69

uania

-0.68

aler

-0.66

enium

-0.64

arling

-0.64

endants

-0.63

urst

-0.62

 doomed

-0.61

md

-0.61

POSITIVE LOGITS

 peoples

0.91

 vain

0.81

 circulation

0.80

escap

0.76

 anybody

0.74

 your

0.73

your

0.72

 YOUR

0.70

 limbo

0.70

our

0.69

Activations Density 0.261%

phrases related to uncertainty or decision-making processes

phrases that express being "in" or "out" of various situations or contexts

No Comments

No Known Activations