INDEX

Explanations

mentions of publications or research findings

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter "M" in varying contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Nep

-0.63

rend

-0.62

 tipped

-0.60

say

-0.60

gets

-0.60

 soup

-0.59

gie

-0.57

 heav

-0.56

 tint

-0.56

 Hath

-0.56

POSITIVE LOGITS

useum

1.20

achine

1.08

ueller

1.05

obile

1.03

umbai

1.02

ISSION

1.01

ortal

1.01

ORPG

1.01

ilitary

0.99

ansion

0.98

Activations Density 0.073%

mentions of publications or research findings

occurrences of the letter "M" in varying contexts

No Comments

No Known Activations