INDEX

Explanations

words related to news events and locations

oai_token-act-pair · gpt-3.5-turbo

events related to injuries or fatalities

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

Newsletter

-1.02

theless

-0.88

↵Âł

-0.87

 ÂłÂł

-0.84

';

-0.83

 Âł Âł Âł Âł Âł Âł Âł Âł

-0.82

:,

-0.79

 UNCLASSIFIED

-0.78

[/

-0.78

-)

-0.77

POSITIVE LOGITS

 watchdog

0.85

 Wednesday

0.68

 Tuesday

0.68

 Thursday

0.65

 Monday

0.64

 TODAY

0.64

 accidentally

0.64

sex

0.64

 Saturday

0.63

 overnight

0.61

Activations Density 1.283%

words related to news events and locations

events related to injuries or fatalities

No Comments

No Known Activations