INDEX

Explanations

dates from the late 1960s and early 1970s

oai_token-act-pair · gpt-3.5-turbo

references to specific years, particularly from the 1960s and early 1970s

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

venge

-0.82

EStreamFrame

-0.69

ndra

-0.67

erate

-0.66

hop

-0.65

 explan

-0.64

plan

-0.64

uler

-0.61

 lawy

-0.61

awed

-0.60

POSITIVE LOGITS

-'

0.85

å¹

0.80

 ILCS

0.76

BCE

0.75

 onwards

0.74

çļ

0.73

 Minutes

0.71

 "$:/

0.68

oslov

0.67

 onward

0.66

Activations Density 0.055%

dates from the late 1960s and early 1970s

references to specific years, particularly from the 1960s and early 1970s

No Comments

No Known Activations