INDEX

Explanations

phrases related to exclusion or limitation

oai_token-act-pair · gpt-3.5-turbo

conjunctions and phrases that indicate conditions or qualifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

oward

-0.88

quished

-0.81

ukong

-0.70

ivating

-0.65

successfully

-0.63

=#

-0.63

 earthqu

-0.63

ãĤ¼ãĤ¦ãĤ¹

-0.62

hur

-0.62

ilaterally

-0.61

POSITIVE LOGITS

 nothing

0.99

 meaningless

0.97

 ignores

0.93

 insignificant

0.92

nothing

0.90

 negligible

0.88

 none

0.87

 nowhere

0.87

 nobody

0.85

 worthless

0.85

Activations Density 0.524%

phrases related to exclusion or limitation

conjunctions and phrases that indicate conditions or qualifications

No Comments

No Known Activations