INDEX

Explanations

the name "Nit" in different contexts and variations

oai_token-act-pair · gpt-3.5-turbo

mentions of nitrogen and related chemical terms

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

wings

-0.87

naire

-0.78

answer

-0.77

 Granger

-0.77

 Jarvis

-0.76

 Weasley

-0.74

arching

-0.73

asley

-0.71

helm

-0.70

 pronoun

-0.68

POSITIVE LOGITS

Nit

3.73

Nit

3.14

nit

2.72

nit

2.71

Ion

1.18

 Amit

1.13

 Neon

1.13

Mit

1.10

Oxy

1.10

 Mush

1.09

Activations Density 0.053%

the name "Nit" in different contexts and variations

mentions of nitrogen and related chemical terms

No Comments

No Known Activations