INDEX

Explanations

references to evolution

oai_token-act-pair · gpt-3.5-turbo

the repetition of the token "ev" at varying activation levels

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Whitman

-0.86

matically

-0.62

 corpor

-0.62

 mature

-0.61

TRY

-0.61

 actionGroup

-0.61

 hitter

-0.61

 common

-0.60

 mail

-0.60

 Bermuda

-0.60

POSITIVE LOGITS

apor

1.15

itability

1.14

olution

1.10

ev

1.10

iev

1.09

olve

1.07

rov

1.07

idently

1.05

ice

1.02

idential

1.01

Activations Density 0.005%

references to evolution

the repetition of the token "ev" at varying activation levels

No Comments

No Known Activations