INDEX

Explanations

occurrences of the word "housed."

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

places or structures where things are located or situated

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

uyomi

-0.72

vious

-0.65

 implication

-0.65

 Blow

-0.63

vous

-0.62

 Carney

-0.61

 parachute

-0.60

 plane

-0.59

sen

-0.59

 Jonas

-0.58

POSITIVE LOGITS

 housed

0.87

owship

0.85

=#

0.81

estinal

0.81

hyde

0.79

ĸļ

0.77

maid

0.74

chool

0.74

ãĤ´ãĥ³

0.74

tern

0.72

Activations Density 0.010%

occurrences of the word "housed."

places or structures where things are located or situated

No Comments

No Known Activations