INDEX

Explanations

words related to discoveries or findings across various contexts

oai_token-act-pair · gpt-3.5-turbo

terms related to discovery and discoveries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

erate

-0.74

osuke

-0.72

oji

-0.69

cussion

-0.67

itent

-0.65

ogun

-0.64

oided

-0.63

attery

-0.63

cour

-0.62

 approved

-0.62

POSITIVE LOGITS

æĥ

0.79

ibility

0.78

Discover

0.75

gdala

0.75

 flaws

0.72

Ô

0.70

 anew

0.70

 Primordial

0.70

Notes

0.70

ledge

0.69

Activations Density 0.046%

words related to discoveries or findings across various contexts

terms related to discovery and discoveries

No Comments

No Known Activations