INDEX

Explanations

references to a specific location, Salt Lake or related terms

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "Salt" and related references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

merce

-1.07

mercial

-0.95

ufact

-0.84

 arrang

-0.83

 therap

-0.81

enegger

-0.78

 livest

-0.69

GROUP

-0.69

umper

-0.67

 WARN

-0.65

POSITIVE LOGITS

zman

1.08

water

0.96

 Salt

0.95

 Lake

0.92

ivas

0.92

itude

0.91

illo

0.91

Salt

0.89

Lake

0.89

ash

0.88

Activations Density 0.007%

references to a specific location, Salt Lake or related terms

occurrences of the word "Salt" and related references

No Comments

No Known Activations