INDEX

Explanations

words related to specific locations, particularly San Francisco in this case

oai_token-act-pair · gpt-3.5-turbo

references to San Francisco

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

ãĥ³ãĤ¸

-0.81

 Cobra

-0.80

ãĥīãĥ©ãĤ´ãĥ³

-0.79

zig

-0.76

ãĥīãĥ©

-0.75

ãĥ¡

-0.75

ãĥĥãĤ¯

-0.73

paio

-0.72

 Spice

-0.72

brand

-0.71

POSITIVE LOGITS

IENT

1.05

ANC

0.86

umbers

0.84

OU

0.81

anova

0.78

atell

0.75

ENN

0.75

hester

0.72

atted

0.72

eton

0.71

Activations Density 0.012%

words related to specific locations, particularly San Francisco in this case

references to San Francisco

No Comments

No Known Activations