INDEX

Explanations

words and phrases related to mental health conditions, particularly depression

oai_token-act-pair · gpt-3.5-turbo

references to mental health issues, particularly depression

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

ateur

-0.79

aque

-0.76

Reply

-0.75

lay

-0.74

quer

-0.73

nel

-0.71

Sources

-0.70

mone

-0.70

join

-0.69

 Foss

-0.69

POSITIVE LOGITS

 relapse

1.14

 symptoms

1.06

 depressive

1.00

 medication

0.99

 depression

0.98

 disorder

0.97

 Symptoms

0.94

 diagnosis

0.94

 symptom

0.90

 medications

0.88

Activations Density 0.044%

words and phrases related to mental health conditions, particularly depression

references to mental health issues, particularly depression

No Comments

No Known Activations