INDEX

Explanations

vague terms or references

oai_token-act-pair · gpt-3.5-turbo

terms related to vagueness or lack of specificity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

mount

-0.75

 Advantage

-0.72

Master

-0.69

 Turn

-0.68

 Bench

-0.67

Enjoy

-0.66

enter

-0.65

issance

-0.64

 Mount

-0.63

 Caption

-0.61

POSITIVE LOGITS

 vague

3.40

 vaguely

1.89

 ambiguous

1.79

neb

1.69

 cryptic

1.60

 fuzzy

1.50

 murky

1.47

 blurry

1.43

 dubious

1.28

 indist

1.26

Activations Density 0.026%

vague terms or references

terms related to vagueness or lack of specificity

No Comments

No Known Activations