INDEX

Explanations

verbs related to giving commands or instructions

oai_token-act-pair · gpt-3.5-turbo

verbs related to taking action or implementing changes

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

rikes

-0.62

 reluct

-0.61

mith

-0.60

ohn

-0.60

lit

-0.60

-0.58

 exert

-0.58

arag

-0.57

idal

-0.57

hell

-0.57

POSITIVE LOGITS

ables

1.25

ings

1.24

 Yourself

1.17

able

1.16

ments

1.14

 yourselves

1.10

 yourself

1.03

INGS

1.01

ability

1.01

acion

0.94

Activations Density 0.383%

verbs related to giving commands or instructions

verbs related to taking action or implementing changes

No Comments

No Known Activations