INDEX

Explanations

references to locations or entities related to 'lin'

oai_token-act-pair · gpt-3.5-turbo

the specific term "lin" as a substring in various contexts

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

awei

-0.92

 Seym

-0.87

ModLoader

-0.77

rahim

-0.76

ansas

-0.74

ADRA

-0.72

tub

-0.70

 mathemat

-0.68

aeper

-0.68

lished

-0.68

POSITIVE LOGITS

ergic

1.04

eman

0.97

ium

0.92

burg

0.91

 Rouge

0.87

strument

0.86

hurst

0.83

fo

0.83

ergy

0.82

ights

0.82

Activations Density 0.016%

references to locations or entities related to 'lin'

the specific term "lin" as a substring in various contexts

No Comments

No Known Activations