INDEX

Explanations

descriptions of appearances or conditions

oai_token-act-pair · gpt-3.5-turbo

occurrences of the phrase "looks like."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

die

-0.74

ricular

-0.72

oled

-0.72

âĹ¼

-0.72

kson

-0.68

sov

-0.68

death

-0.65

renheit

-0.65

verend

-0.64

scribed

-0.64

POSITIVE LOGITS

ynthesis

0.82

":"/

0.78

peed

0.78

paces

0.72

ahead

0.72

metics

0.72

earch

0.72

ynt

0.70

 suspic

0.70

ometimes

0.70

Activations Density 0.021%

descriptions of appearances or conditions

occurrences of the phrase "looks like."

No Comments

No Known Activations