INDEX

Explanations

words related to political actions and decisions

oai_token-act-pair · gpt-3.5-turbo

terms related to political actions and decisions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

verbs

-0.69

 acron

-0.66

 salads

-0.64

 newsletters

-0.58

 fractions

-0.57

 fools

-0.56

ongs

-0.55

 Merchants

-0.55

 stere

-0.55

itudes

-0.55

POSITIVE LOGITS

 someday

1.08

 anytime

1.06

 akin

0.91

 sooner

0.89

 sometime

0.86

 ASAP

0.83

 showdown

0.80

 soon

0.78

ossier

0.78

 anywhere

0.77

Activations Density 0.388%

words related to political actions and decisions

terms related to political actions and decisions

No Comments

No Known Activations