INDEX

Explanations

words related to deference or showing respect for others' opinions or beliefs

oai_token-act-pair · gpt-3.5-turbo

terms related to references

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 proc

-0.69

 captcha

-0.69

 Manufact

-0.65

 pockets

-0.64

 bore

-0.64

 launched

-0.63

 managed

-0.63

 chart

-0.63

 polls

-0.62

band

-0.61

POSITIVE LOGITS

erence

4.53

rence

1.99

erent

1.82

erential

1.70

erences

1.47

iculty

1.40

rences

1.26

inction

1.13

ilde

1.07

cellence

1.01

Activations Density 0.026%

words related to deference or showing respect for others' opinions or beliefs

terms related to references

No Comments

No Known Activations