INDEX

Explanations

phrases containing the word "possible" with a positive association

oai_token-act-pair · gpt-3.5-turbo

instances of the word "possible" used in various contexts related to minimizing or optimizing actions

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

strong

-0.67

riad

-0.61

ĺħ

-0.60

masters

-0.60

Ĭ

-0.59

ĪĴ

-0.59

pse

-0.59

Ħ¢

-0.59

furt

-0.58

rets

-0.58

POSITIVE LOGITS

 without

0.84

 WITHOUT

0.78

 whilst

0.67

 thanks

0.67

 while

0.67

uchin

0.66

without

0.65

 before

0.65

.","

0.62

0.61

Activations Density 0.068%

phrases containing the word "possible" with a positive association

instances of the word "possible" used in various contexts related to minimizing or optimizing actions

No Comments

No Known Activations