INDEX

Explanations

phrases indicating negation or impossibility

oai_token-act-pair · gpt-3.5-turbo

instances of the phrase "couldn't" or variations of it, reflecting inability or impossibility

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 continuum

-0.73

 guiActiveUnfocused

-0.68

ctors

-0.67

 buggy

-0.63

 scene

-0.61

 Golem

-0.60

 condol

-0.59

 Marble

-0.59

 nucleus

-0.59

isode

-0.59

POSITIVE LOGITS

_>

0.90

£

0.87

ı

0.86

âģ

0.85

¹

0.85

âĢķ

0.84

âĢł

0.82

ISIS

0.82

¬

0.81

abo

0.81

Activations Density 0.270%

phrases indicating negation or impossibility

instances of the phrase "couldn't" or variations of it, reflecting inability or impossibility

No Comments

No Known Activations