INDEX

Explanations

names of universities and people, particularly a focus on finding the word "High"

oai_token-act-pair · gpt-3.5-turbo

proper nouns, specifically names and educational institutions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ĸļ

-0.82

 Janeiro

-0.68

schild

-0.67

LOAD

-0.63

PDATE

-0.63

 cuff

-0.63

FACE

-0.62

WAYS

-0.59

 informants

-0.58

umblr

-0.58

POSITIVE LOGITS

sels

0.72

rice

0.68

gall

0.67

oulos

0.66

osph

0.66

estival

0.65

oup

0.63

bourg

0.63

olis

0.61

redible

0.61

Activations Density 0.768%

names of universities and people, particularly a focus on finding the word "High"

proper nouns, specifically names and educational institutions

No Comments

No Known Activations