INDEX

Explanations

phrases related to collaboration or unity

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate collaboration or unity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

lean

-0.69

ãĥĥãĤ¯

-0.68

erved

-0.66

rued

-0.66

Secondly

-0.65

clud

-0.63

erg

-0.61

arted

-0.61

ishment

-0.61

STER

-0.61

POSITIVE LOGITS

 form

1.00

 common

0.91

 mutually

0.89

 mutual

0.86

 create

0.86

 collectively

0.81

 jointly

0.81

 forming

0.80

 devise

0.80

 cohesive

0.80

Activations Density 0.244%

phrases related to collaboration or unity

phrases that indicate collaboration or unity

No Comments

No Known Activations