INDEX

Explanations

terms related to foundational or fundamental concepts

oai_token-act-pair · gpt-3.5-turbo

references to fundamental concepts and basics

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

asus

-0.84

asse

-0.78

gments

-0.74

irection

-0.72

odor

-0.64

ŃĶ

-0.64

vice

-0.64

aton

-0.63

iable

-0.61

atten

-0.60

POSITIVE LOGITS

paces

0.95

chool

0.87

 Reborn

0.85

cape

0.85

heet

0.82

linger

0.80

hooting

0.78

ynthesis

0.78

pace

0.77

poons

0.77

Activations Density 0.052%

terms related to foundational or fundamental concepts

references to fundamental concepts and basics

No Comments

No Known Activations