INDEX

Explanations

names or terms that contain the substring "am" or "AB."

oai_token-act-pair · gpt-3.5-turbo

specific recurring names or identifiers related to characters or subjects in a narrative context

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Fritz

-0.85

 Strauss

-0.84

æł

-0.78

 pengu

-0.78

 Gors

-0.76

 Sloven

-0.75

 Garfield

-0.75

 Slov

-0.74

è¦

-0.72

 Entered

-0.71

POSITIVE LOGITS

am

1.45

AM

1.25

amer

1.24

amia

1.24

ams

1.20

ammy

1.17

im

1.16

oms

1.13

ample

1.13

amn

1.13

Activations Density 0.174%

names or terms that contain the substring "am" or "AB."

specific recurring names or identifiers related to characters or subjects in a narrative context

No Comments

No Known Activations